.''(j(j2
STANTONA. HARRIS,(>RKAI,L) A. HOYACK AND KAHLPOLEERS
therein was the first mixture on which this analytical procedure was tried. The difference of 2.7y0 between observed and calculated percentage of glucose is higher than that encountered in subsequent determinations, and it Wac usually less than 2%. Errors in the di- and trisaccharide fractions tended t o be a little higher (about 230/0). If one of these fractions was small the error was larger than usual, as in the fourth mixture of Table I. Such an error may be counteracted in part by taking a larger sample (see last mixture, Table I). The disadvantage of a large sample ordinarily is the greater time required for the di3tillation. Mixture Containing Fructose.-The misture was composed of fructose hydrate (5.06 g.), glucose (5.00 g.), sucrose (4.99 g . ) , raffinose pentahydrate (5.35 g.). On an anhydrous basis the weights would be 4.60, 5.00, 4.99, 4.54 g.; theoretical percentages 24.1, 26.1, 26.1, 23.7, respectively. Very mild conditions were used to avoid decomposition of the fructose propionate. The degassing operation and the pressure lowering were done a t 140-150 '. Distillation started a t the bath temperature of 165' (0.003 mm.) which was unusually low for mixtures. Ebullition was violent and some fumes were carried away to the trap The material darkened considerably but the distillate was pale yellow in color. Rapid distillation occurred between 168-175 ', after which the temperature was raised gradually t o 195" (0.002 mm.); yield, 11.02 g. in this first fraction. It was believed to contain the fructose pentapropionate and some of the glucose analog. At 195' the distillate became more nearly colorless and the remainder of the glucose pentapropionate was collected separately (8.63 g . ) . After this, there was 10.57 g. of a disaccharide fraction, and 8.37 g. of a weighed residue of trisaccharide The theoretical residue should have been 10.17 g. From the appearance of the fructose fraction as it distilled, some decomposition evidently was occurring. It was estimated that 1.3g. of the 1.8 g. o€ total loss was from the fructose portion. This figure is about 12% of the fructose fraction. It was arrived at otherwise, however, by
Vol. 63
assuming that the loss of about 0.06 g. for each 1 g. of weighed trisaccharide residue, which was noticed in runs without fructose, should maintain itself as the approximate trisaccharide loss if fructose was present. I n this case, it would be 8.4 X 0.06 or 0.5 g. Thus, the combined weight of monosaccharide propionates was 20.95 g. (11.02 8.63 1.3); disaccharide propionate 10.57 g.; and trisaccharide propionate 8.87 g. by difference (10.17 - 1.3). From these data these percentages Follow: monosaccharide 48.9, sucrose 27.3, raffinose 23.8, compared with 50.2, 26.1, 23.7, respectively. If the trisaccharide fraction is taken as 8.37 g. (the weighed residue) then these percentages follow: monosaccharides 49.6, sucrose 27.6, raffinose 22.8. The sample taken contained 24.1 % fructose and 26.1% glucose. The first fraction of distillate (11.02 g. or 12.32 g., corrected) contained glucose as well as fructose. If 1.58 g. is subtracted from the 11.02 g., or if 2.27 g. is taken from the 12.32 g., and added to the weight (8.63 9.) of the second Fraction then the correct ratios for fructose and glucose would appear, The correction of 1.58 g. is 18% of the weight of the second fraction.
+
+
Summary A method of analysis of sugar mixtures is presented which is applicable for mixtures of mono-, di- and trisaccharides. In this method the carbohydrates are converted to propionic esters and distilled under controlled conditions in a special apparatus. The method includes most sugars without the necessity of correction factors and will even include fructose if appropriate corrections are introduced. The accuracy which has been attained for monosaccharides is about 1-2%, and for di- and trisaccharides about 2-4010. EVANSTON, ILLINOIS
RECEIVED JULY 5, 1941
[CONTRIBUTION YROM THE RESEARCH LABORATORIES OF MERCK& eo., INC.]
On the Synthesis of Pantothenic Acid and Derivatives* BY STANTON A. HARRIS, GERALD A. BOYACK AND KARL FOLKERS Subsequent to the announcement of the struc- dium salt of @-alanineto give either ethyl (+)ture and synthesis of pantothenic acid,' details of pantothenate (11) or (+)-pantothenic acid (111). the s t r u c t ~ r eand ~ ~ s~y n t h e ~ i s were ~ * ~ published. The last step in the synthesis involved the reaction between (-) a-hydroxy-P,P-dimethyl-y-butyrolactone (I) and ,&alanine ethyl ester or the so-
*
This paper was presented berore the Organic Division of the American Chemical Society at St. Louis, Missouri on April 8, 1941. (1) Williams and Major, Science, Si, 246 (1940). (2) Mitchell, Weinstock, Jr., Snell, Stanberry and Williams, THIS J O U R N A L , 61, 1776 (1940). (3) Stiller, Keresztesy and Finkelstein, i b i d . , 62, 1779 (1940). (4) Williams, Mitchell, Weinstock and Snell, i b i d . , 62, 1784 (1940). (5) Stiller, Harris, Finkelstein, Keresztesy and Folkers, ihid., 62, 1785 (1940).
I HOCHzC(CH~)~CHOKCONHCH~CHZCOZR' 11. R = H, R' = CzHs IIr. K and R' = H (pantothenic acid) IV. R = CHJCO--, R' = C2H5V. R == ~-NOZC~HICO--,R' = H VI. R = C~HSCH~OCO-,R' C2Hb-
Oct.. 1941
SYNTHESIS OF PANTOTHENIC ACIDAND DERIVATIVES
In the interim, experiments on the reaction of the dl-lactone (I) with @-alanine6and the dllactone with the sodium salt of p-alanine' have been described as yielding pantothenic acid in solution. The reactions between the (-) and dl-lactone and @-alanine methyl ester* and palanine benzyl estersa also have been described. All these investigators have combined the lactone with either a @-alanine ester or sodium salt. This reaction has now been applied successfully to the three diverse acyl derivatives of the lactone to give the corresponding pantothenic acid derivatives. The acetyl derivative of the (-)lactone3 (I) and p-nitrobenzoyl derivative3 were combined with @-alanineethyl ester and the sodium salt of 0-alanine, respectively, to give ethyl monoacetylpantothenate (IV) and mono-p-nitrobenzoylpantothenic acid (V). Of these derivatives, IV was an oil distillable in high vacuum whereas V was crystalline. It was of interest to test mono-p-nitrobenzoylpantothenic acid microbiologically, because nothing is known about derivatives of pantothenic acid having the a-hydroxy group protected. Its activity was kindly determined by Dr. John C. Keresztesy in this Laboratory, who reported that i t was inactive (or had less than 1% of the activity of pantothenic acid) for the growth stimulation of Lactobacillus ca.sei.9 The carbobenzoxy group introducedlO for the protection of amino groups in peptide syntheses has been utilized in these pantothenic acid studies. The carbobenzoxy derivative of the lactone (I) was made and combined with @-alanineester to give ethyl carbobenzoxppantothexiate (VI) which was an oil distillable in a high vacuum. When the carbobenzoxy lactone w2s treated with the sodium salt of @-alanine in aqueous solution, a group interchange took place and N-carbobenzoxy-@alanine" was isolated. A classical method for synthesis of pantothenic acid, which is a substituted N-acyl-@-aminopropionic acid, would be to treat the suitable acyl chloride with a @-alanineester or salt. As soon as the amide character of pantothenic acid was (6) Weinstock, Jr., Arnold, May and Price, Science, 91, 411 (1940). The manner in which @-alaninewas used is not described. (7) Babcock, Jr.. and Jukes, THISJOURNAL, 62, 1628 (1Y40). (8) Reichstein and Grtissner, Help. Ckim. Acta, 23, 1550 (1940). (Sa) Ruhn and Weland, B e . , 73, 971 (1940). (9) Snell, Strong and Peterson, J . Bacl., 38, 203 (1939). (10) Bergmann and Zervas, Bcr., 66, 1192 (1932). (11) Sifferd and du Vigneaud. J . B i d . Chem., 108, 753 (1935).
2663
realized, l2$l3 Woolley, Waisman and Elvehjem13 were successful in applying this method for partial synthesis. Since the acyl group of pantothenic acid was known to possess free hydroxy groups, acetylation was introduced to protect the hydroxy groups prior to the thionyl chloride treatment for preparation of the acyl chloride, After the announcement of the pantothenic acid structure' showing that the lactone moiety was a gamma lactone, the same fundamental acetic anhydridethionyl chloride steps were used by Wo01ley'~J~ for syntheses, and since active pantothenic acid was obtained, it might be assumed that the reaction proceeded by the steps VI1 + VI11 3 I X ---t X or XI -+ 111. When synthetic (-)lactone (I) became available, i t was converted to the sodium salt (VII) and then carried through the above steps.13 Since the final pantothenic acid derivative was apparently a mixture of mono and diacetyl ethyl pantothenate, we investigated the intermediate products of the synthesis with the larger quantities of material which were available to us. These experiments demonstrated that the sodium salt (VII) with acetic anhydride gave the known acetyl lactone (XII) together with the diacetyl acid. Another experiment was made to determine the relative proportion of these two products. The ether extract from the acetylation of the sodium salt was fractionally distilled and i t was found to consist of 73.57, of the acetyl lactone and 17.7% of diacetyl acid. The 73.5% yield of the acetyl lactone is nearly a true value, whereas the 17.7Y0 of diacetyl acid is low because of the distillation loss of this higher boiling fraction. OthersI6 have observed the recovery of lactones on attempting to acylate sodium salts of the hydroxy acids corresponding to lactones. Therefore, we believe that the reactions used by Woolley, et aZ.,l38l41l6for the synthesis of pantothenic acid proceeded only to a small extent in the direction involving interaction of the acid chloride with the &alanine ester or salt and that the major course of the reaction involved the interaction of (12) Williams, Weinstock, Jr., and Mitchell, Abstracts, Division of Organic Chemistry, Amer. Chem. SOC.,Milwaukee, Wis., 1938. (13) Woolley, Waisman and Elvehjem, THIS JOURNAL, 61, 977 (1939); J. B i d . Chem., 129, 673 (1939). (14) Woolley, J . B i d . Ckem., 134, 461 (1940). (15) Woolley, THISJOURNAL, 62, 2281 (1940). (16) For example, Ohle and Wolter [Ber., 63, 843 (1930)l obtained triacetyl-2-ketogluconic acid lactone from the sodium salt in pyridine with acetic anhydride. Dr. E. T. Stiller (unpublished data) in this Laboratory observed a 4milar reaction for other y-hydroxy aliphatic
acids.
2664
STANTON A. HARRIS,GERALD A. BOYACK AND Kmu F O L ~ R S CH: )C-CHOCOC& CHz
1
I
CH2 C=O
\d
,%alanine ester
CHI OH
I I HOCH-C-CH-COzNa I
VOl. 63
IV
CHI OCOCHa
1
(CHaC0)zO
1
A CH&!0OCHz-C-CH-CO2H
I
CH,
CHI
I
VI1
VI11
JSOC4 CHs OCOCHa
CHs OCOCHs
I I palanine ester14 I I CHaCOOCH~-C~CHCONH(CH~)zCOzCzHsfCHaCOOCHz-C-CHCOC1 I
CHI I
1
I CHI
X
I
I
OH-
I11
CHI OCOCHa t
I
OH-
I
CHaCOOCHry-CH-CONH( I
Pantothenic acid
CH:
IX
j3-alanine16 NaOH
$
CHz)zCOzH XI
the acetyl lactone with the p-alanine ester or salt. prepared the diacetyl acid derivative (VIII) When the sodium salt of the acid was treated with indirectly through the following reactions, I + acetic anhydride, freed of excess reagent and then XI11 + XIV + VIII. The diacetyl acid was treated with thionyl chloride (the excess of which then treated with thionyl chloride to give the acid was afterward removed by distillation), a 60% chloride (IX) which in turn was treated with pyield of acetyl lactone was obtained by distilla- alanine ethyl ester to give pure ethyl diacetyltion of the reaction product. In another experi- pantothenate (X) which on saponification yielded ment, after the removal of the excess thionyl chlo- pantothenic acid. ride, the residue conCHs OH CH, OCOCHI tained 4.16% chlorine, liq. I 1 (CHaC0)zO I 1 whereas the theoret- I +HOCHz-C-CH-cO~Hz ---f CHaCOOCHz-C-CHCONH2 3"
I
I
ical amount of chloCHs XI11 CH, XIV rine in the diacetyl CsH11ONO acid chloride is ., V SOClZ 14.15'%. This chlop-alanine IX VI11 rine content indicated ester that the product conReichstein and Griissner8 obtained the racemic sisted of about %yoof the diacetyl acid chloride and about 71y0of acetyl lactone. This content amide by treating the optically active lactones with alcoholic ammonia. Our amide was prepared of acetyl lactone is in good agreement with 73.5% shown to be present in the distillate from the ether with liquid ammonia and was optically active. It was of interest to test OUT samples of ethyl extract described above. The agreement between the acetyl lactone content of the two preparations pantothenate and ethyl monoacetyl pantothenate indicates that reaction between the acetyl lactone microbiologically and in rats and chicks because and thionyl chloride to give a n acid chloride must of previous reports.17 This investigation was done (17) Mitchell, Weinstock, Snell, Stanberry and Williams [THIS have been negligible. JOURNAL, 61, 1776 (1940)]found i t necessary t o hydrolyze methyl Because Of Our experience with low yields in acetylpantothenate for microbiological assay. Snell, Strong and Peterson [Biochcm. J., 31, 1789 (1937)l found that acetylation inacattempts t o acylate the sodium salt of (+)cy,ytivated their growth factor for lactic acid bacteria. Woolley, Waisdihydroxy-p,p-dimethylbutyric acid (VII)? we man, Mickelsen and Elvehjem [ J . B ~ O Z .Chem., isti, 715 (193s)1
-
1
Oct., 1941
SYNTHESIS OF PANTOTHBNIC ACID AND
DERIVATIVES
2665
tone was 19.06 g. (67%) and the yield of the diacetyl acid was 4.56 g. (12%). The main loss in yield occurred in the extraction process. Addition of sodium acetate to the acetylation mixture did not increase the yield of the diacetyl acid. The lactone fraction crystallized when seeded with a known sample of a-acetoxy-B,j3-dimethyl-ybutyrolactone.' After three recrystallizations of the compound from ether, the melting point was raised from 40-41' t o 44-45". It showed no depression of melting point with the above compound; [ a l z 8-13.1' ~ in 95% Ethyl N-( a-Acetoxy-&@-dimethyl-y-hydroxybutyryl)-j3- alcohol (C = 2.72%). aminopropionate (Ethyl Monoacetylpantothenate) 1V.Afial. Calcd. for C&I120+: C, 55.81; H, 7.03. Found: The sodium salt (VII) (2 9.) was acetylated with an excess C, 55.90; H, 7.05. of acetic anhydride as described.'a A crystalline precipiAnother experiment was made with 4.7 g. of sodium tate separated which was filtered, washed with ether and salt for the above acetylation step. The residual oil was found to be sodium acetate. The acetic anhydride was treated with a 100% excess of thionyl chloride, heated a t distilled in vacuo and the residual sodium acetate removed 100' for fifteen minutes, evaporated to remove thionyl by treatment with ether. The oily residue from the ether chloride and distilled at 2 mm. pressure; b. p. 95'. Cryssolution was treated with an excess of thionyl chloride tallization took place on seeding with acetyl lactone. which was removed later by vacuum distillation. The After two crystallizations of the compound from ether, product was condensed with a 20y0 excess of @-alanine the melting point was 41-42', and the mixed melting point ester in pyridine as described.'s There was no visible wa.s 41-43'; yield, 2.83 g. The mother liquor from the evidence of the formation of pyridine hydrochloride. The recrystallization gave a negative test for inorganic chlorine, solution was warmed and the excess of pyridine was re- indicating the absence of a n acid chloride. moved in vucuo. The residue was treated with ice water N-(a-p-Nitrobenzoxy-,9,@-dimethyl-y-hydroxybutyryl)and made acid to congo red paper with hydrochloric acid. 8-aminopropionic Acid (Mononitrobenzoate of PantoThe oily droplets were extracted by chloroform which was thenic Acid) V.-An attempt was made t o prepare the washed, dried and concentrated. The oily concentrate dinitrobenzoate of a-y-dihydroxy-@,P-dimethylbutyric acid was distilled a t 10-6-10-6 mm. a t a bath temperature of by esterifying the sodium salt in pyridine with p-nitro100-110'. After redistillation, the following analysis was benzoyl chloride. However, the product was the monoobtained. nitrobenzoate described in a previous paper.6 This ester Anal. Calcd. for C13H2306N: C, 53.96; H , 8.01; N, (0.62 g.) was heated with 0.233 g. of the sodium salt of 4.84; acetyl, 14.8%. Found: C, 54.22, 53.96; H, 8.30, &alanine for one hour a t 100". After a few minutes, the 8.30; N, 4.68; acetyl, 19.75. mixture became gummy and when cooled, solidified to a Anal. Calcd. for C I ~ H ~ ~ OC, ~N 54.37; : H, 7.61; N, brittle mass. After the material was dissolved in 10 cc. 4.23; acetyl, 25.98. of water, a few crystals of starting material separated. The analyses indicated that the substance was a mixture When the filtrate was acidified with hydrochloric acid, a of the mono- and diacetates of ethyl pantothenate. This gummy precipitate separated. It readily dissolved in was confirmed by boiling the substance for four hours with alcohol and reprecipitated as an oil on the addition of 0.1 N sodium hydroxide solution when only 2.3 moles was water. On standing overnight, the oil crystallized, and it used for the saponification. was recrystallized from acetone by the addition of water, Pure monoacetyl ethyl pantothenate was made by m. p. 137-138". I t was the mononitrobenzoate of pantoheating acetyl lactone and &alanine ester together with- thenic acid; [ a I z 9+4.5' ~ in 95% alcohol ( C = 0.78%). out solvent. The resulting oil was distilled a t 10-6 mm. Anal. Calcd. for C16H~~N208:C, 52.17; H, 5.47; N, A non-viscous oil was obtained below loo', and a viscous 7.61. Found: C, 52.23; H, 5.64; N, 7.38. oil was obtained a t about 120" bath temperature. This a-Carbobenzdioxy+,@-dimethyl-y-butyro1actone.-An oil was hygroscopic. attempt was made to prepare the dicarbobenzoxy ester of Anal. Calcd. for C13H230~N: C, 53.96; H, 8.01; N, a ,y-dihydroxy-@,@-dimethylbutyric acid by treating an 4.84; acetyl, 14.88. Found: C, 53.83; H, 7.86; N, 4.63; alkaline solution of the acid with carbobenzoxy chloride. acetyl, 13.2. Little or no reaction was observable. On acidification of Acetylation of Sodium cu,-y-Dihydroxy-P1j3-dimethyl- the alkaline solution, no insoluble acid was obtained. The monocarbobenzoxy lactone was made as follows. butyrate.-The sodium salt (28.2 g.) was refluxed for onehalf hour with acetic anhydride and the solution concen- A benzene solution containing 6 g. of the lactone and trated to dryness. After adding 20 cc. each of 2.5 N 8.65 g. (1 equivalent) of antipyrine was added to a benzene hydrochloric acid and water, an oily, ether-soluble layer solution containing 4.6 g. (1 equivalent) of phosgene. separated. The mixture was extracted for six hours with Crystalline antipyrine hydrochloride separated. After ether; the extract was concentrated and the sirupy residue fifteen minutes, a benzene solution containing equivalent was distilled a t mm. The yield of monoacetyl lac- quantities of benzyl alcohol (5 g.) and antipyrine (8.65 8.) was added. Heat was evolved and additional antipyrine reported that acetylation inactivated their chick antidermatitis hydrochloride separated. The mixture was heated a t factor and that the activity was recovered on hydrolysis. Griissaer, Giitzi-Fichter and Reichsteia [Hclo. Chim. Acto, 23,1276 (1940)J 100' for fifteen minutes and filtered. The filtrate was washed three times with water, and then dried over reported that ethyl pantothenate was active in rats.
in the Merck Institute for Therapeutic Research by Dr. Unna and Mr. Mushett who found that these esters of pantothenic acid were active in rats and chicks, but were practically inactive microbiologically. Their studies on the biological utilization of these esters will be described elsewhere. Experimental Part
2666
STANTON A. HARRIS,GERALD A. BOYACK AND KARLFOLKERS
Vol. 63
calcium chloride. After distillation of solvent, the sirup ethanol; +5.7', C = 2.3% in ethyl acetate; -5.4', C = crystallized. Recrystallization was done by solution in 1.5y0in dioxane. alcohol and addition of water to turbidity. On cooling, Anal. Calcd. for CloH1705N:C, 51.93; H, 7.41; N, the ester crystallized; m. p. 78'; yield, 5.05 g., 41.8%; 6.06. Found: C, 51.66; H , 7.34; N, 5.90. [ a I 2 *+12.3" ~ in 93% alcohol (c = 2.1%). a,y-Diacetoxy-@,,%dimethylbutyric Acid (VIII).-A soAnal. Calcd. for ClrHlsOs: C, 63.63; H, 6.09. Found: lution of 11.54 g. of the diacetoxyamide (XIV) in about 50 C, 63.48; H, 6.32. cc. of glacial acetic acid was treated with 20 cc. of amyl N-Carbobenzoxy-6-alanine from the Condensation of nitrite and heated a t 100" for eighty minutes. The soluthe a-Carbobenzdioxy-p,p-dimethyl-y-butyrolactonewith tion was concentrated in vacuo and the residue taken up in the Sodium Salt of b-Alanine.-A mixture of 1.32 g. water in which it dissolved very slowly. After making (0.005 mole) of the carbobenzoxylactone and 0.005 mole just alkaline with sodium hydroxide and extracting with of the sodium salt of &alanine was heated for two hours chloroform, the solution was acidified to pH 2 and exa t 100' with frequent stirring. On the addition of water, tracted with chloroform. This solution was concentrated 0.78 g. of unchanged lactone crystallized. When the to a sirup under 2 mm. pressure. The residue (7.52 g.) water solution was acidified, shiny crystals separated titrated for 90% acid, and was distilled a t 10-8 mm. a t which were recrystallized from alcohol and water; m. p. 100' bath temperature. Three fractions were taken, each 103'. Judged by analysis and melting point, the com- of which contained nitrogen by qualitative test. The pound was the carbobenzoxy derivative of b-alanine third fraction was repurified as described above and then which was originally prepared by Sifferd and du Vigneaud.lo redistilled. I t was free of nitrogen; [ a ] 2 6 D -2.6', C = Anal. Calcd. for C I I H ~ ~ N OC,~ 59.19; : H, 5.83; N, 1.512 in methanol; 0 in ether. 6.28. Found: C, 58.81; H, 5.79; N, 6.20. Anal. Calcd. for CloHleOe: C, 51.72; H, 6.95; acetyl, Ethyl N-(a-Carbobenzdioxy-@,@-dimethyl--,-hydroxy- 37.06. Found: C, 51.60; H , 7.20; acetyl, 33.93. butyryl)-j3-aminopropionate.-After 1.9 g. of freshly disEthyl N-( a-y-Diacetoxy-p,p-dimethylbutyry1)-p-aminotilled p-alanine ester was mixed with 4.2 g. of carbobenzoxy propionate. (Ethyl Diacetylpantothenate) (X.)-A mixlactone and heated a t 100' for one and one-half hours, it ture of 4.4 g. of the diacetyl acid and 2.77 cc. of thionyl was shaken with water and extracted with ether. The chloride was heated a t 100' for one-half hour and the ether solution was washed with dilute hydrochloric acid excess of thionyl chloride was removed under vacuum. and water, dried over calcium chloride, filtered with char- The residue was dissolved in 5 cc. of pyridine containing coal and distilled. The oil was distilled between 140-150' 2.23 g. of @-alanineethyl ester. Heat was evolved and bath temperature a t 4 X 10-8 mm. pressure. It was the the solution became slightly colored. After standing for carbobenzoxy ester of ethyl pantothenate (VI). seventy-two hours, the excess pyridine was removed by Anal. Calcd. for C I B H ~ ~ N OC, I : 59.83; H, 7.14; N, distillation under vacuum, the sirup was dissolved in chloroform which was extracted first with dilute hydro3.67. Found: C, 59.57; H, 6.86; N, 3.89. chloric acid, then with water and sodium bicarbonate a--pDihydroxy-p,p-dimethylbutyramide (XIII).-The lactone, on treatment with either aqueous ammonia or al- solution, and finally dried over calcium chloride. After coholic ammonia, followed by evaporation of the solvent, solvent removal, the sirup was distilled a t 10-8 mm. and yielded a substance which was crystallized from alcohol; 110-120' bath temperature. The viscous sirup was hygroni. p. 135-136'. Analyses indicated that it was mainly scopic and care was taken nct to expose it to moist air the ammonium salt of a,y-dihydroxy-b,fi-dimethylbutyricbefore analysis; yield, 3.06 g. (48.5%); [ a ] 2 K D +24.2', C = 2.08Y0 in ether. acid. Sublimation under vacuum yielded the lactone. Anal. Calcd. for C15H2507N: C, 54.37; H, 7.61; N, Anal. Calcd. for CeHljOdN: C, 43.62; H, 9.15; N, 4.23; acetyl, 25.98. Found: C, 54.18; H, 7.33; N, 4.04; 8.48. Found: C, 43.96; H , 8.73; N, 7.64. The amide (XII) was made by enclosing 13.6 g. of the acetyl, 24.04. A solution of 0.11 g. of the ester was dissolved in 10 cc. (-) lactone and 50 cc. of liquid ammonia in a bomb tube, allowing the mixture to warm to 25' (the lactone dissolved) of 0.5 N barium hydroxide and allowed to stand a t 20-25' and t o stand overnight. After opening the cooled tube for one hour. The barium ion was removed as the sulfate and allowing the ammonia t o evaporate, the amide crystal- and the solution brought to pH 7.5 with sodium hydroxide. lized. The last of the ammonia was removed under Aliquots were assayed microbiologicallys and were found vacuum and the residue recrystallized from ethyl acetate; t o contain 50y0 of the theoretical amount of pantothenic yield, 15.13 g., 98.5%; m. p. 92-94'; [a]29D $30.9' in acid. HzO (C = 2.09%). Acknowledgment.-We greatly appreciate the Anal. Calcd. for CeHlaOaN: C, 48.97; H, 8.90; N, cooperation of Messrs. D. F. Hayman, W. Reiss 9.52. Found: C, 48.93; H.8.90; N, 9.52. and H. S. Clark in performing the microanalyses. a,yDiacetoxy-j3,p-dimethylbutyrdde (XIV).-The amide was acetylated with acetic anhydride both with and Summary without pyridine. The main fraction came over at 125' bath temperature a t 10-6 mm. I t analyzed for a,-pdiThe acetyl and carbobenzoxy derivatives of acetoxy-&P-dimethylbutyramide. The specific rotation (-)a hydroxy /3,/3 d;methyl y - butyrovaried considerably depending on the solvent: [ C Y ] % lactone were treated with p-alanine ethyl ester, +6.8', C = 5.3% inether; -0.7', C = 2.8% inchloroform; -10.3', C = 3.7y0 in water; -3.2', C = 2.2% in absolute and the p-nitrobenzoyl lactone with the sodium
-
-
-
-
DISSOCIATION CONSTANT OF BROMCRESOL GREENI N WATER
Oct., 1941
salt of 8-alanine, to give the corresponding ethyl acetylpantothenate, ethyl carbobenzoxypantothenate and P-nitrobenzoylpantothenic acid. p Nitrobenzoylpantothenic acid was inactive in microbiological tests. Ethyl pantothenate and ethyl acetylpantothenate were practically inactive in microbiological tests but were active in rats and chicks. Evidence is described which shows that when sodium CY,y-dihydroxy-@,@-dimethylbutyrate is acetylated, then treated with thionyl chloride, then treated with 8-alanine or its ester, the major mechanism of the formation of the pantothenic
[CONTRIBUTION FROM
THE
2667
acid derivative involves the condensation of a-acetoxy-@,@-dimethylbutyrolactone with 8alanine, while only a minor part of the synthesis is accomplished through the condensation of the diacetyl acid chloride with @-alanine. Pure CY,y-diacetoxy-@,fl-dimetliylbutyric acid was made by the treatment of a,y-diacetoxy@,P-dimethylbutyramide with amyl nitrite in acetic acid. Subsequent treatment with thionyl chloride and @-alanine ester gave pure ethyl diacetylpantothenate which was saponified to pantothenic acid. RAHWAY, NEWJERSEY
RECEIVED JUNE 17, 1941
DEPARTMENT OF CHEMISTRY AND CHEMICAL ENGINEERING OF THE UNIVERSITY OF PENNSYLVANIA]
The Dissociation Constant of Brom Cresol Green in Water‘ BY MARTINKILPATRICK
__
Several determinations have been made of the equilibrium constant for the reaction Brom cresol green (yellow)
+ H20
H80”
+ Brom cresol green (blue)
TABLE I
THEEQUILIBRIUM CONSTANT AT 25’ green -k Acetate
Brom (yellow)
Acetic acid
(1)
and the dissociation constant K c = C H ~ O ~ ~ ( A ,=. HC~HO~)OC+B ~ I C A ,
(2)
a t zero ion concentration has been reported by various w o r k e r ~ . ~ ~Since ~ ~ ~ the ~ 5 values vary from 1.03 t o 1.20 X i t has seemed worthwhile to present a summary of a number of determinations made in this Laboratory. The first series involves the colorimetric determination of the equilibrium constant, K A i B , of the reaction
+
Brom cresol green Acetate (yellow) Acetic acid
+ Brom cresol green (blue)
(3)
where K A ~= B
CB CAi C A CBI
(4)
The constant KkB may be related to the thermodynamic equilibrium constant [&B] by the equation log K l i ~= log [KA,B]’f A(Z.4
- Z ~ i ) f -ki BAL (5)
where p = ‘/z x C Z a and Z, and Z,, are the charges on the buffer acid and indicator acid, re(1) Aided by a grant from the Penrose Fund of the American Philosophical Society. (2) Sendroy and Hnstingr, J . B i d Chcm., 82, 197 (1920). (3) Guggenheim and Schindler. J . Phyr. Chem., 88, 543 (1934). (4) Kilpatrick, Chem. Rev., 16, 57 (1935). (6) Minnick a i d Kilpntrick, J. Phrr. Cbcm.. CS, 259 (1939).
+ Brom cresol green (blue)
Ionic strength, P
0.00079 .00082 .00129 .00162 .0018a .00202 .00208 .00216 .00233 .00246 .00256 .00302 .00375 .00506 .00516 .00602
.00622 ,00740 .00784 ,00950 .01000 .01000 .01035 .01152 , 0 1286 ,01477 .01542 .02000 to2020
log
KA~B,
-obsd.
-1.827
-1.826 -1.833 -1.837 -1.844 -1.847
-1.851 1.851
-1.847 -1.849 -1.859 -1.849
-1.855 -1.860 -1.861 -1.866 -1.855 -1.868 -1.867 -1.873
1.886 i.877 i-. 8 8 6 -1.887 -1.897 -1.881 -1.889 1.895 j.897